Circulant Quantum Channels

• Circulant quantum channels are completely positive trace-preserving maps exhibiting cyclic symmetry with explicit links to circulant matrix algebras and representation theory.
• Their structure allows precise characterization of quantum capacities, including zero-error capacity quantified via the Lovász theta number and semidefinite programming.
• They serve as analytical models in quantum transport, decoherence, and dissipative dynamics, offering clear insights into quantum thermodynamics and resource theories.

Circulant quantum channels are completely positive trace-preserving (CPTP) maps on finite-dimensional Hilbert spaces, distinguished by their invariance under cyclic group actions and their intimate relationship with circulant matrix algebras, representation theory, and quantum information measures. This class encompasses several central families—mixed-unitary channels generated by cyclic shifts, Weyl-diagonal channels, and block-circulant quantum Markov semigroups—each admitting explicit characterization due to the underlying group symmetry. Their applications range from quantum zero-error capacity and coherence theory to explicit models of decoherence, transport, and entropy production in open quantum systems.

1. Algebraic and Structural Definitions

A canonical circulant quantum channel $\Phi$ on $\mathbb{C}^n$ is defined by a convex combination of cyclic shifts:

$$\Phi(\rho) = \sum_{k=0}^{n-1} p_k\, U^k\, \rho\, (U^k)^\dagger\,, \qquad U = \sum_{i=1}^n |i\oplus1\rangle\langle i|$$

where $p_k \ge 0$, $\sum_k p_k = 1$, and $\oplus$ denotes addition mod $n$ (Xie et al., 23 Jan 2026). The Kraus representation is

$$K_k = \sqrt{p_k}\, U^k, \qquad \sum_{k=0}^{n-1} K_k^\dagger K_k = I\,.$$

A matrix $X\in \mathcal{M}_n(\mathbb{C})$ is circulant if $X_{i,j} = c_{j\ominus i}$ for some vector $(c_0,\ldots,c_{n-1})$. Every circulant matrix commutes with each $U^k$. The image of $\Phi$ coincides exactly with the set of circulant matrices, and $\Phi$ is idempotent ($\Phi^2 = \Phi$) and unital.

For multipartite systems, the class of Weyl-diagonal channels generalizes circulant symmetry: for $\mathcal{H} = (\mathbb{C}^d)^{\otimes N}$,

$$\Phi(\rho) = \frac{1}{d^N} \sum_{\vec m, \vec n} \lambda_{\vec m, \vec n}\, W_{\vec m, \vec n}\; \rho\; W_{\vec m, \vec n}^\dagger$$

where $W_{m,n} = X^m Z^n$ are the single-qudit Weyl operators and the coefficients $\lambda$ encode the channel parameters; complete positivity and trace-preservation impose specific Fourier-analytic constraints (Basile et al., 2023).

2. Zero-Error Capacity and Circulant Graph-Induced Channels

In a foundational framework, the action of a circulant quantum channel can be derived from an orthonormal representation (OOR) of a circulant graph $G = X(\mathbb{Z}_n, C)$, with $C$ specified by (equal-sized) cyclotomic cosets. The OOR is generated via a "shift" unitary $U$ and the eigenbasis of the adjacency matrix $A_G$. The induced classical-quantum (CQ) channel maps $$|k\rangle \langle k| \mapsto |u_k\rangle \langle u_k|$$, with $|u_k\rangle$ obtained from $U^k$ acting on an initial vector dependent on the spectrum of $A_G$ (Lai et al., 2015).

The one-shot zero-error capacity $C_{0,NS}^{(1)}(N)$, for a CQ channel $N$ assisted by quantum nonsignalling correlations, is given by

$$C_{0,NS}^{(1)}(N) = \log\left\lfloor \Upsilon(N) \right\rfloor$$

where $\Upsilon(N)$ is the optimal value of a semidefinite program (SDP) depending only on the Kraus operators. For these circulant CQ channels, $\Upsilon(N)$ equals the Lovász $\vartheta(G)$ number of the underlying graph. Thus, the asymptotic zero-error capacity is

$C_{0,NS}(N) = \log \vartheta(G)$

imparting an operational meaning to the Lovász number as the ultimate classical-quantum zero-error capacity under maximally permissive (nonsignalling) assistance.

Table: Zero-Error Capacity for Selected Circulant Graphs

Graph Type	$\vartheta(G)$	Asymptotic Capacity $C_{0,NS}(N)$
Cycle $C_n$ (odd $n$)	$\frac{n\cos(\pi/n)}{1+\cos(\pi/n)}$	$\log \frac{n\cos(\pi/n)}{1+\cos(\pi/n)}$
Paley $QR_p$ ($p\equiv 1\pmod{4}$)	$\sqrt p$	$\frac{1}{2}\log p$
Cubic-residue $CR_p$ ($p\equiv 1\pmod{3}$)	see main text	$\log \vartheta(G)$ (no closed form)

These results apply for graphs including cycle, Paley, and cubic-residue graphs arising from cyclotomic coset constructions (Lai et al., 2015).

3. Circulant Quantum Markov Semigroups and Dissipation

A circulant quantum Markov semigroup (qms) on $\mathcal{H} = \mathbb{C}^p \otimes \mathbb{C}^q$ is generated by a Gorini–Kossakowski–Sudarshan–Lindblad (GKSL) form with Lindblad operators $L_{i,j} = \sqrt{a(i,j)}\;(J_p^i \otimes J_q^j)$, $a:\mathbb{Z}_p\times\mathbb{Z}_q\to[0,\infty)$, $a(0,0) = 0$, and $J_p, J_q$ the primary shift unitaries (Bolaños-Servín et al., 2012).

The generator $L(X) = \Phi(X) - X$, with

$$\Phi(X) = \sum_{i,j} a(i,j)\,(J_p^i\otimes J_q^j)^* X (J_p^i\otimes J_q^j)$$

is block-circulant with circulant blocks. This structure enables explicit diagonalization and a cycle decomposition directly analogous to Markov chains. The entropy production rate (QEPR) for the stationary state $\rho = I/(pq)$ is

$$\sigma(\rho) = \frac{1}{2pq} \sum_{(i,j)} \left[a(i,j) - a(p-i, q-j)\right] \log \frac{a(i,j)}{a(p-i, q-j)}$$

The vanishing of QEPR characterizes quantum detailed balance ($a(i,j) = a(p-i, q-j)$ for all $(i,j)$), while positive QEPR quantifies nonequilibrium dissipation.

A plausible implication is that circulant qms, due to their analytical tractability, serve as minimal models for quantum transport, heat engines, and decoherence phenomena exhibiting periodic symmetries.

4. Weyl-Diagonal and Component-Erasing Channels

Weyl-diagonal channels, defined on $(\mathbb{C}^d)^{\otimes N}$ as

$$\Phi(\rho) = \frac{1}{d^N} \sum_{\vec m, \vec n} \lambda_{\vec m, \vec n}\, W_{\vec m, \vec n}\; \rho\; W_{\vec m, \vec n}^\dagger$$

admit full characterization via their Choi matrices, which are block-circulant and diagonalized by the discrete Fourier transform. Complete positivity requires the Fourier transforms of the $\lambda$ coefficients be non-negative (Basile et al., 2023). The set of such channels is convex, and its extreme points are precisely the conjugations by Weyl operators.

A subclass, "component-erasing channels" (Editor's term), erase all but a specified subgroup of Weyl components, and are fully classified by additive subgroups of the label group $G=\mathbb{Z}_d^N\oplus\mathbb{Z}_d^N$ and associated group homomorphisms. The set of component-erasing channels forms a semigroup under composition, and minimal generating sets correspond to codimension-one subgroups with linear characters.

5. Entanglement-Breaking Property and Operational Implications

Among all mixed-unitary circulant channels $\Phi_{\boldsymbol{p}}(\cdot)$, only the uniform twirl ($p_k = 1/n$) is entanglement-breaking (Xie et al., 23 Jan 2026). Its Choi matrix is separable and, in low dimensions, positive under partial transpose (PPT). The uniform circulant channel projects any input to a convex combination of basis projectors, completely erasing quantum correlations.

In quantum resource theories, $\Phi$ is an incoherent operation in the computational basis, and the $\ell_p$-norm coherence of a state satisfies

$C_p(\rho)\; \ge\; C_p(\Phi(\rho))$

with explicit closed-form expressions for the output. Further, the circulant channel attains extremal values of high-order Bargmann invariants, connecting matrix symmetries to phases in quantum interference.

6. Connections to Perfect State Transfer and Circulant Spin Networks

Circulant graphs arise in spin network models supporting perfect state transfer (PST). For a weighted circulant on $n$ vertices, PST exists if and only if $n$ is even and special conditions on the weights hold, reducible to a statement about integrality of the eigenvalues: all eigenvalues must be integers, and odd weights must occur at divisors $n/2$ or $n/4$, all others divisible by $4$ (Bašić, 2011). This spectral rigidity reflects the high symmetry and algebraic structure underlying circulant quantum channels, constraining dynamical and information-theoretic capacities.

7. Physical Interpretations and Future Perspectives

Circulant quantum channels serve as prototypical models in quantum information theory—both as idealizations capturing maximal symmetry (e.g., uniform mixing, decoherence, symmetry reduction) and as analytically tractable systems for exploring fundamental limits (such as zero-error capacity or entropy production). Their explicit connection to the Lovász theta number provides a precise graph-theoretic measure of quantum channel capacity under optimal correlations, yielding direct operational meaning to previously abstract parameters (Lai et al., 2015).

A plausible implication is that further generalization to inhomogeneous or higher-dimensional cyclic group actions, as well as their interplay with local and global invariants, will yield deeper insight into robust quantum communication and error correction in symmetric environments, with potential applications to networked quantum systems, dissipative engineering, and quantum thermodynamics.